Bone metastasis is common in patients with prostate cancer and is a frequent source of morbidity, including bone pain or fracture. Assessment of skeletal bone metastases in these patients has been with imaging modalities. However, bone scintigraphy is not highly specific and fails to detect areas of bone degradation, a feature common to bony metastases. In addition, many patients with a clinical response to therapy characterized by a declining PSA and improved well-being may not have an immediate corresponding improvement in the bone scintigraphy scan. Thus, biochemical markers of bone metabolism in blood have been explored as indicators of bone turnover for their potential as prognostic and/or predictive variables. As part of a randomized NCI-sponsored phase II trial of a matrix metalloproteinase inhibitor in patients with hormone refractory prostate cancer with skeletal metastases, our group prospectively evaluated serum markers of bone turnover from 69 patients and correlated the results with patient outcome. We found that several baseline markers of bone formation (osteocalcin, procollagen N-terminal propeptides: PINP &PIIINP, total alkaline phosphates) and resorption (N-telopeptide, pyridinoline, deoxypyridinoline) in serum had statistically significant prognostic value. Data from trials employing the endothelin-A antagonist Atrasentan in prostate cancer patients showed that levels of total and bone alkaline phosphates significantly predicted time to progression. We will prospectively validate these encouraging preliminary results in a similar, larger, patient cohort through the recently initiated randomized phase III Southwest Oncology Group trial (S0421) of docetaxel/prednisone with or without the endothelin-A antagonist Atrasentan (n=706) in patients with metastatic hormone refractory prostate cancer. The hypothesis is that baseline levels of bone metabolism markers in sera collected from patients with metastatic hormone refractory prostate cancer will be of prognostic value, while serial levels of these same markers will predict enhanced response and survival following therapy with a bone-targeted agent such as Atrasentan. This hypothesis will be tested through the following specific aims: 1) measure levels of selected markers of bone formation (procollagen I amino-terminal propeptides: PINP, total and bone alkaline phosphates) and resorption (N- telopeptide, deoxypyridinoline) in sera collected at baseline and serially from patients enrolled in SWOG 0421, a randomized phase III trial of docetaxel/prednisone with or without Atrasentan in patients with hormone refractory prostate cancer with skeletal metastases;2) correlate the results of these marker studies with tumor response, progression-free survival, and overall survival to detect prognostic (baseline or pre-treatment sera) and/or predictive (serial sera) value;3) identify prognostic groups based on baseline bone markers and other clinical and disease-related factors, and to evaluate whether change in a bone marker is a surrogate for survival in this setting.